There is growing evidence that plant and animal species are arranged in hierarchies of relative competitive performance. More work is needed to determine which plant traits best predict relative competitive performance. We therefore measured relative competitive performance of 63 terrestrial herbaceous plant species using Trichostema brachiatum as a reference species (that is, phytometer or target species). The neighbour species came from a wide array of terrestrial vegetation types (e.g. rock barrens, alvars, old fields), and represented a wide array of growth forms (e.g. small rosette species such as Saxifraga virginiensis and large clonal graminoids such as Agropyron repens). The experiment was repeated with two pot sizes: large (control) and small (stress treatment). Relative competitive performance in large pots (controls) was highly correlated with that in small pots (stress treatment) (r = 0.90, p < 0.001). The hierarchy of relative competitive performance in the large pots was also highly correlated with the hierarchy in the small (stressed) pots (r s = 0.91, p < 0.001). Principal components analysis and multiple linear regression showed that plant size (measured by total biomass, above-ground biomass, below-ground biomass, canopy area, height and leaf area index) and leaf shape (measured as length to width ratio, length, width) were the two characteristics that best predicted relative competitive performance (large pots, r 2 = 0.55; small pots, r 2 = 0.48).

A small coryneform bacterium was consistently isolated from sugarcane with ratoon stunting disease and shown to be the causal agent. A similar bacterium was isolated from Bermuda grass. Both strains multiplied in sugarcane and Bermuda grass, but the Bermuda grass strain did not incite the symptoms of ratoon stunting disease in sugarcane. Shoot growth in Bermuda grass was retarded by both strains.

No abstract

No abstract